Predicting Clostridioides difficile infection outcomes with explainable machine learning.

BACKGROUND: Clostridioides difficile infection results in life-threatening short-term outcomes and the potential for subsequent recurrent infection. Predicting these outcomes at diagnosis, when important clinical decisions need to be made, has proven to be a difficult task. METHODS: 52 clinical features from existing models or the literature were collected retrospectively within ±48 h of diagnosis among 1660 inpatient infections. A modified desirability of outcome ranking (DOOR) was designed to encompass clinically-important severe events attributable to the acute infection (intensive care transfer due to sepsis, shock, colectomy/ileostomy, mortality) and/or 60-day recurrence. A deep neural network was constructed and interpreted using SHapley Additive exPlanations (SHAP). High-importance features were used to train a reduced, shallow network and performance was compared to existing conventional models (7 severity, 7 recurrence; after summing DOOR probabilities to align with conventional binary outputs) using area under the ROC curve (AUROC) and DeLong tests. FINDINGS: The full (52-feature) model achieved an out-of-sample AUROC 0.823 for severity and 0.678 for recurrence. SHAP identified 13 unique, highly-important features (age, hypotension, initial treatment, onset, PCR cycle threshold, number of prior episodes, antibiotic exposure, fever, hypotension, pressors, leukocytosis, creatinine, lactate) that were used to train a reduced model, which performed similarly to the full model (severity AUROC difference P = 0.130; recurrence P = 0.426) and significantly better than the top severity model (reduced model predicting severity 0.837, ATLAS 0.749; P = 0.001). The reduced model also outperformed the top recurrence model, but this was not statistically-significant (reduced model recurrence AUROC 0.653, IDSA Recurrence Risk Criteria 0.595; P = 0.196). The final, reduced model was deployed as a web application with real-time SHAP explanations. INTERPRETATION: Our final model outperformed existing severity and recurrence models; however, it requires external validation. A DOOR output allows specific clinical questions to be asked with explainable predictions that can be feasibly implemented with limited computing resources. FUNDING: National Institutes of Health-Institute of Allergy and Infectious Diseases.

Validation of clinical risk tools for recurrent Clostridioides difficile infection.

OBJECTIVE: We sought to validate available tools for predicting recurrent C. difficile infection (CDI) including recurrence risk scores (by Larrainzar-Coghen, Reveles, D'Agostino, Cobo, and Eyre et al) alongside consensus guidelines risk criteria, the leading severity score (ATLAS), and PCR cycle threshold (as marker of fecal organism burden) using electronic medical records. DESIGN: Retrospective cohort study validating previously described tools. SETTING: Tertiary care academic hospital. PATIENTS: Hospitalized adult patients with CDI at University of Virginia Medical Center. METHODS: Risk scores were calculated within ±48 hours of index CDI diagnosis using a large retrospective cohort of 1,519 inpatient infections spanning 7 years and compared using area under the receiver operating characteristic curve (AUROC) and the DeLong test. Recurrent CDI events (defined as a repeat positive test or symptom relapse within 60 days requiring retreatment) were confirmed by clinician chart review. RESULTS: Reveles et al tool achieved the highest AUROC of 0.523 (and 0.537 among a subcohort of 1,230 patients with their first occurrence of CDI), which was not substantially better than other tools including the current IDSA/SHEA C. difficile guidelines or PCR cycle threshold (AUROC: 0.564), regardless of prior infection history. CONCLUSIONS: All tools performed poorly for predicting recurrent C. difficile infection (AUROC range: 0.488-0.564), especially among patients with a prior history of infection (AUROC range: 0.436-0.591). Future studies may benefit from considering novel biomarkers and/or higher-dimensional models that could augment or replace existing tools that underperform.

Reclassification of Pre-Heart Failure Stages Using Cardiac Biomarkers: The ARIC Study.

BACKGROUND: The recent heart failure (HF) guideline recommends the inclusion of cardiac biomarkers in defining Stage B HF. OBJECTIVES: The authors evaluated the impact of incorporating cardiac biomarkers to reclassify HF in 5,324 participants (mean age: 75.8 years) without prevalent HF enrolled in the ARIC (Atherosclerosis Risk In Communities) study and assessed prognosis of Stage B using cardiac biomarkers. METHODS: Using N-terminal pro-B-type natriuretic peptide (<125 pg/mL or ≥125 pg/mL), high-sensitivity troponin T (<14 ng/L or ≥14 ng/L), and abnormal cardiac structure/function by echocardiography, individuals were classified as Stage Anew and Stage Bnew HF, respectively. Stage Bnew was further evaluated as elevated biomarker only, abnormal echocardiogram only, and abnormalities in both (echo + biomarker). The authors assessed risk for incident HF and all-cause death using Cox regression. RESULTS: Overall, 4,326 (81.3%) individuals were classified as Stage Bnew with 1,123 (21.1%) meeting criteria for elevated biomarkers only. Compared with Stage Anew, Stage Bnew was associated with increased risk for incident HF (HR: 3.70 [95% CI: 2.58-5.30]) and death (HR: 1.94 [95% CI: 1.53-2.46]). Stage Bbiomarkers only and Stage Becho only were associated with increased HF risk, whereas Stage Bbiomarkers only was also associated with increased death. Stage Becho+biomarker had the highest risk for HF (HR: 6.34 [95% CI: 4.37-9.19]) and death (HR: 2.53 [95% CI: 1.98-3.23]). CONCLUSIONS: Incorporating biomarkers based on the new HF guideline reclassified approximately 1 in 5 older adults without prevalent HF to Stage B. The routine measurement of biomarkers can help to identify individuals at higher HF risk who may benefit most from HF prevention efforts.

Impact of dementia on marriage: a qualitative systematic review.

This qualitative review explored the impact of dementia on marriage. The method was informed by systematic review and qualitative research methodologies. A comprehensive search of major databases was undertaken. The search identified 115 studies on the topic; 23 met the inclusion criteria and 19 were appraised as being of good methodological quality. Two major themes and five sub-themes emerged from the analysis of included studies, transition and loss. The theme of transition encompassed the three sub-themes of relationship, roles and intimacy and reflected the changes in the marriage and marital relationship that accompanied dementia. The theme of loss encompassed the two sub-themes of loss of a partner and loss of a marriage, which reflected the many losses that accompanied dementia. However, transition and loss were inter-related, because each change was accompanied by loss and each loss produced another change in the marriage.

Respite services for older people.

Respite is an important strategy for supporting family carers and care recipients. However, it is a complex intervention and the term 'respite' is often used as a generic descriptor for a diverse service. The aim of this study was to explore the characteristics of respite services for older people in South Australia. A 32 item questionnaire that was developed from the literature was distributed to coordinators of respite services for older people. The findings show that respite is most often a pre-planned service offered to clients for a week or more. It encompasses many different types; the most common was residential care facility overnight respite. Although there is flexibility in the service, there is a scope to increase the choices that are available to carers and care recipients.